Opened 10 years ago

Closed 10 years ago

#4463 closed Bug (duplicate)

download from webseed sometimes stops before finishing

Reported by: T_X Owned by: jordan
Priority: Normal Milestone: None Set
Component: libtransmission Version: 2.33+
Severity: Normal Keywords: webseed
Cc:

Description

Ok, I thought it'd be working now after fixing #4437 / changeset r12740, but sometimes the torrent still doesn't finish. I deleted and readded the torrent from #4437 in the GUI 10 times, 5 of them finished successfully, the other ones stalled after a couple of percent.

Somehow the URL still seems broken sometimes. I've added some debug output (see attachments) and it looks like transmission is sometimes using the URL 'http://x-realis.dyndns.org/jamendo3/45452' for a file, with the actual path to the file in the URL missing.

PS: The server is not doing any redirections this time (compared to #4445)

PPS: If it helps for debugging I could probably set up a webseed on a university server instead of behind a residential internet uplink.

Attachments (6)

Lorenzo's Music - BAZAR -- Jamendo - OGG Vorbis q3 - 2011.08.26 [www.jamendo.com]-webseed-not-redirected.torrent (3.6 KB) - added by T_X 10 years ago.
Torrent file used for the tests
test.sh (802 bytes) - added by T_X 10 years ago.
Scripted test procedure
0001-debugging-stuff.patch (4.4 KB) - added by T_X 10 years ago.
Patch for extra debug output
logfile (10.2 KB) - added by T_X 10 years ago.
Logfile with extra debug output
0001-Revert-trunk-libT-4437-Multi-file-webseeds-don-t-wor.patch (2.5 KB) - added by T_X 10 years ago.
0002-libtransmission-Fix-webseeding-issue-for-multi-file-.patch (2.5 KB) - added by T_X 10 years ago.

Download all attachments as: .zip

Change History (9)

Changed 10 years ago by T_X

Scripted test procedure

Changed 10 years ago by T_X

Patch for extra debug output

Changed 10 years ago by T_X

Logfile with extra debug output

comment:1 Changed 10 years ago by T_X

With r12740, you are setting t->bad_range here, but you are freeing 't' three lines later with t->bad_range's set value never being used?

Could this be the reason (t->bad_range being read from an uninitialized value here and here) why it sometimes works and sometimes not?

comment:2 Changed 10 years ago by jordan

Thanks for the patches, T_X. My commit in r12740 was pretty bad. :|

Reopening #4437 and closing this ticket as a duplicate.

comment:3 Changed 10 years ago by jordan

  • Resolution set to duplicate
  • Status changed from new to closed
Note: See TracTickets for help on using tickets.